feature request: setting to negate instantPowerW value readings

[bennetyee]

Since the panel hardware appears to allow the current transformers to be installed backwards and installers often do so, a firmware fix to simply negate the reported readings on a per-circuit basis would remove the need to fix the hardware. (Based on discussions seen, there's no safety reason to turn the CTs around.)

Existing software should ignore an extra reversedCtClamps field in the json response payload, and get the corrected value for the instantPowerW attribute.

[bennetyee]

sorry if this is not the right place for this idea -- are there any SPAN engineers here? could somebody mention it to the firmware team?

[dcj]

Thanks for the suggestion, but I think the premise here doesn't quite apply to SPAN hardware. SPAN panels don't use installer-placed CT clamps — current sensing is integrated into the panel itself, with sensors built into the busbar structure at each branch breaker position. There's no clamp to orient and no installer step where the CT could be put on backwards, so the "fix the hardware-in-software" path that the proposal is aimed at isn't really an option people need.

Sign convention on instantPowerW is also deliberate and consistent: it's defined from the panel's perspective, so positive = power flowing into the panel from that boundary (e.g. a circuit backfeeding because it has a PV inverter on it), and negative = power flowing out of the panel (normal load consumption). I'll be expanding the public docs to spell this out fully — the short version is "positive = imported direction, negative = exported direction, where imported/exported are defined per context." A per-circuit negate flag would let one panel report the opposite sign convention from every other panel for the same physical situation, which would break every downstream consumer that interprets sign meaningfully — Home Assistant energy dashboards, utility export accounting, anything that distinguishes import from export.

If you're seeing readings that look "wrong" on a specific circuit, it's almost certainly one of: (a) a generation circuit (PV, battery, EV V2H) showing legitimately positive numbers because power is flowing back into the panel rather than out of it, or (b) the import/export sign convention not being documented well enough yet — which is what #3 is about. Happy to dig in if you can share which circuit and what you were expecting vs. what you're seeing.

Going to close this one as not-planned, but please reopen or file a fresh issue if there's a specific reading that doesn't match expectations and isn't covered by the above.

---

Edited 2026-05-27: corrected the sign-convention direction (the panel-perspective rule is positive = imported direction, not negative as I initially wrote). The argument against a per-circuit negate flag is unchanged.

[bennetyee]

oooh! negative means load. i googled for why the value was negative and was led down the rabbit hole of reversed CT. the docs are indeed confusing, esp since the phone app doesn't show negative signs.

thanks!