Is the battery charging or discharging?

Read the convention at the source

Current direction is represented by a convention. A device protocol may assign one sign to charging while another assigns that sign to discharging. Software can normalize the device’s raw value so the dashboard uses a consistent convention.

Sunstead’s internal battery-power convention is positive for charging and negative for discharging. Some supported inverter firmware needs a documented, identity-specific polarity correction during ingestion. That is why the exact model, firmware and protocol matter when reporting an apparent reversal.

Do not apply a blanket sign flip to every device because one screen looks wrong. A correct BMS reading and an incorrectly interpreted inverter reading can coexist.

Use power as a cross-check, not a new authority

The example assumes the voltage and current refer to the same boundary at the same time. It does not include every conversion loss between the battery and AC loads. A software-derived power value is not an independent confirmation if it was calculated from the same possibly misinterpreted current.

Collect an ordinary change

Watch a normal transition in the household or solar conditions. Record the inverter’s own charging/discharging indication, the monitor value, the timestamp and battery SOC trend. Use more than one sample; SOC may update slowly and should not be expected to move visibly every second.

A clear independent charging indication alongside consistently labeled discharge is useful evidence. “Solar is larger than load” is weaker evidence when grid flow, other chargers or some loads are unmeasured. Avoid using an incomplete energy balance as the only judge.

Separate source disagreement from stale data

Compare measurements from the same time. A battery can legitimately switch direction between polls. A screen frozen on an older sample can seem to contradict a fresh one without a polarity bug.

Check whether the values describe one pack or the whole bank. A charging controller reports its output, which is not necessarily the bank’s net current after all other loads. The SmartShunt operation manual explains the importance of following current into and out of the monitored battery; the measurement boundary is essential.

Report enough information to reproduce the interpretation

Save the exact device model and firmware, connection type, field labels, sign convention shown, time zone and several observations with their independent indications. Include whether the value comes directly from a BMS or from inverter terminals.

A fix may require a parser correction for a specific identity rather than a user setting. Keep historical data provenance in mind: correcting a future sign interpretation does not necessarily rewrite old aggregate energy records. Ask which intervals are affected before using earlier charge/discharge totals in a comparison.